Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
Review
. 2009 Jun;30(6):1857-65.
doi: 10.1002/hbm.20745.

Source connectivity analysis with MEG and EEG

Jan-Mathijs Schoffelen  1 , Joachim Gross
Affiliations
Review

Source connectivity analysis with MEG and EEG

Jan-Mathijs Schoffelen et al. Hum Brain Mapp. 2009 Jun.

Abstract

Interactions between functionally specialized brain regions are crucial for normal brain function. Magnetoencephalography (MEG) and electroencephalography (EEG) are techniques suited to capture these interactions, because they provide whole head measurements of brain activity in the millisecond range. More than one sensor picks up the activity of an underlying source. This field spread severely limits the utility of connectivity measures computed directly between sensor recordings. Consequentially, neuronal interactions should be studied on the level of the reconstructed sources. This article reviews several methods that have been applied to investigate interactions between brain regions in source space. We will mainly focus on the different measures used to quantify connectivity, and on the different strategies adopted to identify regions of interest. Despite various successful accounts of MEG and EEG source connectivity, caution with respect to the interpretation of the results is still warranted. This is due to the fact that effects of field spread can never be completely abolished in source space. However, in this very exciting and developing field of research this cautionary note should not discourage researchers from further investigation into the connectivity between neuronal sources.

PubMed Disclaimer

Figures

Figure 1
Figure 1
The effects of field spread confound sensor level estimates of connectivity measures. (A) Schematic representation of how the activity of a neuronal source is picked up by different sensors. (B) The absolute value of the correlation coefficient between all pairs of measured signals as a function of magnetometer distance. (C) Topographical representation of the correlation with respect to a reference sensor (highlighted in white).
Figure 2
Figure 2
An estimated increase in connectivity may have different causes. The left column in each panel represents the sources, which are arranged in three compartments. Each circle represents a source, and these comprise the actual neural sources in the central “brain” compartment, and the irrelevant noise sources in the upper and lower compartments. The neural sources share a common component (a degree of connectivity), which is depicted as the black part of the circles. The circles in the middle column represent two sensor signals. Each of the sensors picks up a different mixture of the underlying sources. The partitioning of the sensor signals represents the relative contributions of the respective sources. The size of the circles represents the amplitude of the signal picked up at these sensors. The right column in each panel represents in the total area of the partitioned circle the degree of connectivity estimated between the two sensor signals. This degree of connectivity results from the total area of overlap between the colored partitions in the sensor signals. (A1, A2) Modulation of estimated connectivity due to a change in connectivity between neuronal sources of interest. The scenario A1 shows 4 sources (four circles in leftmost column). The different colors represent different uncorrelated signal components. The time series of the two sources in the middle contain a common component that makes up a specific part of their individual signal. Both sources are picked up by two sensors (circles in the middle column). Because of the noise sources only part of the sensor signals consists of signals of neural origin. This part is further subdivided into the different relative contributions of the two sources. Because of the fact that the upper sensor is closer to the upper source than to the lower source, it “sees” more of the upper source. Therefore, the yellow partition in the upper sensor is bigger than the red partition. The black part of the sensor signals represents the common component projected from the sources and has in fact been “diluted” by the noise sources. The degree of connectivity is derived by assessing the total overlap of colored partitions between the two sensors. Thus, it consists of the true common component picked up by the sensors (in black) and of the overlap in the colored partitions. In scenario A2, the common component in the activity of the neural sources is increased. Despite the diluting effect of the noise sources, an increase in the degree of connectivity can still be detected. (B1, B2) Modulation of estimated connectivity due to an increase in amplitude of a single neuronal source. (C1, C2) Modulation of estimated connectivity due to a decrease in amplitude of noise sources. (D1, D2) Modulation of estimated connectivity due to an increase in amplitude of the neuronal sources of interest, with a concurrent decrease of connectivity between them.
Figure 3
Figure 3
Effects of field spread are not totally abolished in source space. (A, B) Absolute value of the correlation coefficient as a function of dipole distance between the estimated activity at all pairs of 1,000 dipole locations on the cortical sheet (black), and between the estimated activity at all pairs of 10 of these locations at which activity was simulated (red). Two different inverse methods were used: an LCMV‐beamformer (A) and a minimum norm solution (B). The dipole orientations were unconstrained in both approaches. No regularization was applied. (C, D) The estimated difference in correlation between two “conditions” of simulated data, in which the signal‐to‐noise ratio was equal across conditions. Same conventions as in (A, B). (E, F) The estimated difference in correlation between two “conditions” of simulated data, in which the signal‐to‐noise ratio was about 30% lower in the second condition. Same conventions as in (A, B).
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Astolfi L,Cincotti F,Mattia D,Salinari S,Babiloni C,Basilisco A,Rossini PM,Ding L,Ni Y,He B,Marciani MG,Babiloni F ( 2004): Estimation of the effective and functional human cortical connectivity with structural equation modeling and directed transfer function applied to high‐resolution EEG. Magn Reson Imaging 22: 1457–1470. - PubMed
    1. Astolfi L,Cincotti F,Mattia D,Babiloni C,Carducci F,Basilisco A,Rossini PM,Salinari S,Ding L,Ni Y,He B,Babiloni F ( 2005): Assessing cortical functional connectivity by linear inverse estimation and directed transfer function: simulations and application to real data. Clin Neurophysiol 116: 920–932. - PubMed
    1. Astolfi L,Cincotti F,Mattia D,De Vico Fallani F,Tocci A,Colosimo A,Salinari S,Marciani MG,Hesse W,Witte H,Ursino M,Zavaglia M,Babiloni F ( 2008): Tracking the time‐varying cortical connectivity patterns by adaptive multivariate estimators. IEEE Trans Biomed Eng 55: 902–913. - PubMed
    1. Babiloni F,Cincotti F,Babiloni C,Carducci F,Mattia D,Astolfi L,Basilisco A,Rossini PM,Ding L,Ni Y,Cheng J,Christine K,Sweeney J,He B ( 2005): Estimation of the cortical functional connectivity with the multimodal integration of high‐resolution EEG and fMRI data by directed transfer function. Neuroimage 24: 118–131. - PubMed
    1. Baccala LA,Sameshima K ( 2001): Partial directed coherence: A new concept in neural structure determination. Biol Cybern 84: 463–474. - PubMed
  NODES
Note 2
Project 1
twitter 2